Search

This report is the edited transcript of a seminar given at Lincoln Laboratory on 2 March 1965. The cyclicity of interest in millimeter wave research is traced from the time of Hertz, when investigations of this part of the electromagnetic spectrum began. The waxing and waning of millimeter wave research are traced as exciting new fields are discovered which diverted the interest of physicists. Each re-emergence of millimeter wave research has been more robust, often because the results of the diversions were important to improved techniques. Early millimeter wave apparatus is described. The narrative of the development of a millimeter wave radar for the detection of the moon relates the opening of this spectral region to radar astronomy. Other applications are mentioned, along with future possibilities. A chronology, a list of large millimeter wave antennas, and a bibliography of review papers are included.

A technique is presented for calculating the radar cross section of a conducting cone-sphere as a function of cone angle, aspect angle, tip radius, base radius, and wavelength for both cw and short-pulse incident signals. A modification of the physical optics approximation is used in which the incident sinusoidal wave is replaced by a pulse of triangular waves, permitting the physical optics integrals to be replaced by finite summations. Creeping waves are included by using the known results for a sphere combined with an aspect- angle dependence obtained from a ray-tracing model. For long incident pulses, the calculated cross sections show good agreement with cw measured values. For short incident pulses, the behavior of individual scattering centers may be studied as a function of the target parameters.

The study is concerned with the identification of certain target characteristics through the analysis of received signals. Signal phase and amplitude as functions of time as well as the frequency spectra characteristics relating to discrete and distributed targets are investigated. The fundamental aspects of information received by a pulsed radar are reexamined with particular emphasis on the possibilities of extracting signature criteria from the signal phase as a function of the radar. Particular significance is attached to the stationarity of the phase for isolated point-targets, compared to the phase variations in the pulse associated with multiple targets entering and leaving the pulse pocket. Coherent phase discontinuities are analyzed in connection with regular distributions of point sources.

The Source Distribution Technique ('SDT') embodies a comparatively simple and unrestricted computational method applicable to the complex problem of scattering from arbitrarily shaped geometries. The procedure exhibits a greater degree of flexibility and accuracy than present day approximation techniques. It is a numerical procedure for determination of the radar cross section for complex targets, based on an integral equation approach. It can be employed to determine the electromagnetic scattering from arbitrarily shaped two-and-threedimensional geometries in the difficult resonant frequency ranges in which classical solutions are not available. Combined with the Norair Physical Optics procedure the 'SDT' can be employed to determine the electromagnetic scattering of a variety of shapes, including large scattering objects; i.e. > 2 lambda. Verification of the technique was performed on the Norair radar reflectivity range. An image-plane was utilized to measure the current distribution and scattered fields. A brief outline of the basic procedure is described together with a few examples to demonstrate its validity.

Measurements of power spectra and distribution functions of radar echo fluctuation have been made on PB4Y and SNB airplanes. Data for angle of arrival and amplitude of the echo were obtained from the power spectra. The performance of a monopulse radar, under certain conditions, is superior to that of a sequentially-lobed radar. (Author)

A doppler-sense separation technique, applicable to the Madre radar system, separates radar targets into approaching and receding targets. The presentation of unambiguous doppler frequencies of up to one-half the radar sampling rate is made possible. In addition, three desirable collateral features can be obtained: (a) some suppression of earth backscatter returns, (b) simple i-f bandwidth control, and (c) the availability of range-gated and essentially continuous-wave signals. The separation technique has a theoretical signal-to-noise ratio improvement of 3 db.

Two smooth polished 14-in.-diameter aluminum spheres weighing 2.2 lb and 22 lb have been simultaneously injected into nearly identical 600-naut-mi circular polar orbits. These spheres will provide (a) standard radar targets and (b) observations of the relative change in the orbital parameters with time. Since the scattering cross section of these spheres is less than 0.1 sq ft at 108 Mc, they are not adequate for Space Surveillance System calibration at the present time, although they will be seen more often when system frequency is increased to 216 Mc and transmitter power is increased. They are seen often enough at present, however, so that good orbital elements are obtained from the U.S. Navy Space Surveillance System observations, and it has been possible to identify the light satellite and the heavy one from the relative change in semimajor axis caused by residual atmospheric drag. A density measurement has been obtained from the decrement of the semimajor axis of the light sphere that agrees to within ten percent with the accepted value for the quiet-sun period.

This is the third in a series of summary reports on the general subject of electromagnetic and acoustical scattering by certain bodies of simple shape. Consideration is limited in general to problems of diffraction or scattering where the source of energy is exterior to the scatterer (one exception is the case of a point dipole located at the tip of a spheroid, which is immediately obtainable from a more general form). The body of the report consists of three distinct components, the first and most extensive consisting of a largely verbal discussion of the methods and principles employed in the various solutions of spheroidal functions, the second being a tabulation of the most essential results and the third containing the graphical representations of these and the experimental findings.

An approach to the construction of a radar employing the 120-foot Haystack antenna at a wavelength of approximately 4 cm., a transmitter capable of producing 500 Kw of average power, and a maser preamplifier for the receiver is described. The system is expected to have a threshold (signal-to-noise in a 1 - cps filter) of 355 decibels. Extended integration can increase this factor by about 10 decibels. This radar performance will permit measurements on the inner planets of the solar system throughout their orbital period. Measurements on Jupiter can also be attempted, and a possible detection of Saturn is implied. Experiments exploiting this performance are described, including the conduct of a 4th test of general relativity, the precision determination of the orbits of the inner planets, a measurement of the astronomical unit with increased accuracy, better determination of planetary masses and radii, and measurements on planetary surfaces. These measurements are not only of interest to astronomers and cosmologists but also of critical importance to the accurate guidance and navigation of vehicles in space. The roles played by the (L-band) Millstone radar tracker, and the (UHF) ionospheric radar in current space research are also briefly discussed.

The report studies the electromagnetic radar scattering parameters for a complex airborne target at high frequencies. The components of the target are modeled by simple geometrical shapes including ellipsoids, elliptical plates, ogives, and semi-individual component are derived in terms of simple analytical functions. These are subsequently combined, in a consistent manner that takes into consideration the geometrical phase delay and shadowing effect, to yield the total scattered field from the target. Extensive numerical results are presented for the T-33 aircraft and the BQM-34A drone. A comparison of our theoretical computations with experimental data measured by a nanosecond short- pulsed radar reveals reasonably good agreement.

This report states the mathematics and describes the computer program used to convert inertial celestial coordinates to radar pointing angles, and vice versa, in the Haystack Antenna Pointing System.

The purpose of the report is to present a number of results regarding errors in radar measurements produced by propagation through the ionosphere. The emphasis is on the dependence of these errors on radar frequency, and on the possibility of determining and removing the errors by making simultaneous measurements at two radar frequencies. The ionospheric effects discussed are those on radar range, elevation angle, coherent doppler velocity, pulse shape, and wave polarization.

A computer program, called the Planetary Ephemeris Program (PEP), is being written at Lincoln Laboratory. The purpose of the program is to improve planetary and lunar ephemerides using the results of radar and optical observations. In this report, the author derives the differential equations that are numerically integrated in PEP to determine as functions of time the positions and velocities of the planets, of the Earth-Moon barycenter and of the Moon, and the partial derivatives of these positions and velocities with respect to initial conditions, masses and other parameters. Newtonian theory with the usual unrigorous general relativistic corrections is employed. The equations of motion and the equations for the partial derivatives with respect to initial conditions are presented in the form needed in the Encke's method of integration used in PEP.

Studies of the electron-density, electron and ion temperatures in the F-region were made by means of ground-based radar observations at the Millstone Hill Radar Observatory. A 70-meter parabolic antenna directed vertically and a 2.5-Mw pulse radar operating at 440 Mcps were employed for these measurements which were conducted for periods of 30 hours at approximately weekly intervals throughout 1963. Examination of the echo power as a function of height leads to a profile of electron density with height, provided the electron and ion temperatures are the same (T sub e = T sub i). Additional measurements of the spectra of the signals corresponding to different heights permit the ratio T sub e/T sub i to be determined and, where this is different from unity, the observed profile can then be corrected for the effect on the scattering introduced by the inequality in temperature. Results of observations extending over a period of one year from February 1963 to January 1964 are presented. The ratio T sub e/T sub i achieved a maximum value approximately 2.0 to 2.6 at a height of about 300 km soon after down, irrespective of the season. There was little change in height dependence in this ratio throughout the daylight hours, and at sunset the ratio fell with a time constant of the order of an hour. At night T sub e/T sub i was occasionally close to unity, but more often a significant difference remained in the temperatures at all heights.

During the fall of 1967, radar was employed to study the scattering properties of the planet Venus at 3.8-cm wavelength. An increase in the transmitter power to 300 kW CW and a reduction in the system noise temperature to 60 degrees K provided a considerable improvement in the radar performance. The frequency power spectra of the echoes were determined by digital Fourier analysis of the received signals. The total power in each spectrum was computed to yield the value for the radar cross section of Venus. These average 1.75 percent of the projected area of the disk.

Bistatic scattering cross sections of a thin, perfectly conducting cylinder are developed using the variational method. This method was used by Tai (Stanford Research Institute, Tech. Rept. no. 18, March 1951) for obtaining monostatic (back scatter) radar cross sections. Resulting expressions are developed for wires of 3/2 WL/O length or less, where WL/O is the free space wavelength. The wires are assumed to be in air or vacuo. The resulting calculations apply to the case where the incident electric vector lies in the incident plane and the receiving antenna is polarized parallel to the scattering plane.

Electric instrumentation was developed for air-to-air propagation on measurements at 250-, 1000-, and 3000-mc ranges. A square-wave-modulated crystal-controlled transmitter for 245 and 1000 mc was used with an 18-c- bandwidth filter-amplifier and an APR-4 receiver, while a peak reader with an internally generated carrier was used with the receiver for short-pulse modulation measurements at 256, 2880, and 3295 mc. The rectified output from either the filter-amplifier or peak reader operated a recorder which continuously recorded the received signal on an approximately logarithmic scale. The receiving system using square-wave modulation detected and recorded 107-dbm signals and had a dynamic range of well over 100 db. The short-pulse receiving system detected and recorded signals down to the noise level of the receiver, and had a dynamic range of over 85 db. The 2 systems operated satisfactory with existing radio and radar equipment.

A series of firings was executed in the 100-ft Range K of the VKF for the purpose of determining the causes of radar reflections from the wakes of hypervelocity spheres. A 35-kmc, focused, oblique radar was employed as the primary instrumentation. In addition, receiving and parasite antennas were used to measure transmission and specular reflection. The results of these experiments indicate that, when using the 35-kmc microwave equipment described here, a detectable radar reflection in most cases was obtained from the wake of a hypervelocity sphere (10,000 to 27,000 ft/sec in these experiments) only if the sphere ablated. Measurements of turbulent wake velocity were shown to agree with the predicted values. The region of transition from laminar to turbulent flow was also defined.

Scatter field and radar response of an infinitely long straight metallic wire are derived assuming that the wire is illuminated by an antenna radiating a Gaussian beam of narrow beam width. The scatter field has essentially the same distribution in any plane through the wire axis; it varies from plane to plane by only an amplitude factor. The radar response is at a maximum for vertical incidence - when the beam axis intersects the wire axis at right angles - and decreases exponentially with increasing inclination of the beam axis against this direction. No side lobes are obtained for Gaussian illumination, at least not in the interesting range of small angular deviations from vertical incidence where the radar response has appreciable values. Two cases can be distinguished: (1) the wire crosses the Fresnel region of the antenna and (2) the wire is situated in the antenna far zone. In the former case, the theory is limited to wire radii not exceeding the beam radius at the antenna; in the latter case, arbitrary wire radii are admissible. If the wire is located in the far field region of the antenna, the expression derived for the radar response can be generalized so that it applies to any antenna characteristic. The generalized expression shows that a radar cross section can be assigned to the wire even though it has been assumed to be infinitely long; this radar cross section increases linearly with distance between antenna and wire.

The design and performance of the two radars, which operate simultaneously through a wire-grid beam splitter, are described in detail. A doppler signal simulator provides absolute calibration of each radar so that the dynamic radar cross section of the projectile in flight can be measured to within =one decibel.

The question whether the measurement of radar cross-sections at different frequencies provides a useful diagnostic tool for ascertaining the electron density distribution of spherically symmetric plasma clouds is investigated. This is accomplished by comparing the calculated radar cross- sections of characteristic plasmas with increasing and decreasing refractive index. Exact analytical expressions for the radar crosssections of several typical plasma spheres with increasing and decreasing refractive index are calculated. The calculations are based on an exact wave treatment of the scattering problem. Part I of the present report contains the exact analytical results obtained, while Part II will be devoted to the numerical evaluation of these results, as well as to asymptotic expressions for the limiting cases of high and low frequencies.

The Millstone Hill Thomson (incoherent) scatter radar system has been operated routinely since 1963 to perform a synoptic study of F-region electron densities, and electron and ion temperatures. This report describes system changes made in 1968 which considerably increased the accuracy of the measurements and allowed their extension to higher altitudes. These changes have also made it possible to measure the vertical velocity of the plasma over the altitude range 450 to 900 km to an accuracy on the order of 5 to 10 m/sec, depending upon altitude and time of day. Of even greater significance, complete machine reduction of the results is now possible so that considerable savings in time and effort have been secured in analyzing the data. The new system permits all the radar data to be gathered in the digital computer in real time, therby eliminating the need for post real-time processing of magnetic-tape recordings of the signals. Furthermore, it is now possible to transmit the data to other workers in computer-usable form. This report describes the main functions of the computer program required to analyze the measurements, and lists the times of all measurements made with the new system prior to 1 January 1970. Examples of these results are presented and discussed.

The equations of motion of a body under the influence of the earth's gravity and atmospheric drag are obtained in radar polar coordinates. The choice of this coordinate system may have important advantages in various tracking and filtering algorithms. A rotating ellipsoidal earth with gravitational terms up through the second (easily extendable) harmonic is assumed. The effects of a non-spherical earth along with the corresponding effects of the higher gravitational harmonics can be important in the estimation of ballistic coefficients at very low deceleration levels.

The instrument is altogether not similar to conventional range finders - it rather resembles a radar set. The basic difference here lies in the fact that instead of radio waves the range finder uses a beam of light.

As part of an investigation of the radar backscatter properties of axisymmetric conducting shapes, a study has been made of backscattering by a cone-sphere. This body is comprised of a circular cone with a 25 degree included cone angle, terminated by a tangent spherical segment. Measurements of backscatter cross sections over all aspect angles are presented for wavelengths, in the range from one-tenth to twice the radius of the spherical segment. Comparison of measured values with theoretical predictions shows that satisfactory agreement is obtained for all aspect angles.

Millimeter-wave radar systems was analyzed to determine their low- angle tracking capability. The narrow beam widths that can be obtained at these frequencies with antennas of reasonable size permit a target to be resolved in angle from its image. Because of the severe effects of tropospheric attenuation, the study was limited to two frequency bands, 35 and 94 GHz, which lie near the principal minima of the attenuation curve. Theoretical expressions for radar tracking accuracy, together with existing attenuation data, have been used to compare the performance of realizable 35-GHz (K sub a-band) and 94-GHz (W-band) radars with that of an existing 5280-MHz (C-band) radar. An antenna diameter of 12 feet was assumed for all three systems. It is found that the tracking range of the millimeter-wave radars is limited severely by atmospheric effects and by state-of-the-art transmitters and receivers. But there is a significant increase, as compared with the C-band radar, in the extent of the low angle region within which precision (0.1 mil) tracking can be realized.

The radar for computing the signal-to-noise ratio of radar echoes from the sun is outlined in this report. The radial distribution of electron density in the corona, the coronal temperature, solar noise, galactic noise, and radar-system parameters are taken into account. Examples are given of computations for two solar radar systems that will soon be in operation.

The laminar wake of a re-entry vehicle is simulated by a system of cylindrical plasma shells. The radar cross-section per unit length is found, taking into account the energy losses due to collisions. Numerical results await completion of a computer program.

An earlier theoretical study of radar reflection from a rough planetary surface is extended to include the case where the surface correlation function consists of two or more components. When both large- and small-scale structures are simultaneously present, it is found that the latter may completely dominate the autocorrelation function of the echo and thus render the former undetectable by c-w methods. An additional finding is that the large-scale structure may be detectable in the angular power spectrum obtained from very short pulses as a separate pip at the origin. Experimental confirmation for the lunar case is described. The effective radar gain of the surface derived by the methods of physical optics is found to have a maximum value of unity which leads to a minimum value of 3 for the dielectric constant of the surface. The spectrum of the surface fluctuations inferred from the wavelength-dependence of the surface slope is found to have a gap for components having a scale of the order of a few meters.

Free-flight measurements of the flow about both non-ablating and ablating spheres, sphere-capped cylinders, and cones are reported. Included are radiometer measurements in the 0.2-5.5 micron band, image converter and schlieren photographs, microwave measurements of wake ionization, headon radar cross-section measurements, backscatter radar cross-section measurements of turbulent trails, and measurements of precursor ionization. Radiation results for non-ablating spheres, the effect of contaminants, and a comparison between sphere and cone results are presented. The status of the head-on radar results which include anomalous absorption is discussed. Precursor ionization ahead of non-ablating spheres was identified and attributed to photoionization. Comparative results of both schlieren and radar systems, which involve transition from laminar to turbulent flow and scattering from turbulent wakes are outlined. For cones, conditions cover velocities up to 22,000 ft/sec and pressures up to 150 mm Hg, while for spheres, velocities range up to 24,000 ft/ sec and pressures up to 300 mm Hg.

The concept of the polarization scattering matrix is presented in this report. Measurement techniques leading to the determination of the scattering matrix on a radar range are discussed. The polarization dependence of measurements on a ground plane range are noted. Matrix-determination techniques are examined in the context of groundplane-range operation. The utility of performing scattering matrix measurements on the RAT SCAT range is discussed. Finally a specific measurement technique is chosen for the scattering matrix solution; consideration of operational ease and measurement accuracy forms the basis of this choice.

The fluctuating torques applied by wind forces are an important source of error in the servo systems required to aim large antennas for radar and radio astronomy. Direct measurements were made of the wind-induced torque acting about the elevation axis of a 60-ft Kennedy paraboloidal antenna. Torque was measured by a strain-gage bridge and recorded on magnetic tape. Sixty runs, each of 33 min duration, were recorded. A mean(over) squared torque spectral- density curve was obtained for each run. A two-variable function of the form used by Newton and others gives a good fit with these data. The half-power frequency values ranged from 0.029 to 0.29 radians per second, with a median value of 0.11 radian per second. A second corner frequency in the vicinity of 2. 0 radians per second was also observed. Peak torques were observed during certain runs, and equivalent aerodynamic moment coefficients were derived. Root-mean-square values of wind-torque variation were obtained for all runs.